DDavid's Notes
EN中
Paper Trade library

Experiment 01

Best Company of the Year

A transparent annual momentum experiment inspired by this question: what happens when we buy the prior calendar year's strongest available S&P 500 company, then hold it for one full year?

Historical research record
This is a concentrated, hindsight-sensitive historical experiment, not a recommendation to chase last year's winner. It owns one company at a time and can experience losses far larger than a diversified fund.

Twenty-cohort result

Same contributions, three ending values

Each column receives the same US$1,000 at the beginning of each of the twenty entry years, then carries every contribution to the end of 2025. The only difference is what each envelope owns.

Selected-company ladder

US$201,413

SPY ladder

US$91,245

QQQ ladder

US$166,906

These are historical study totals, not forecasts. The selected-company result is much more concentrated than either ETF benchmark and includes three clearly labeled corporate-action reconstructions.

Experiment 01

How the three retired tickers were carried forward

  1. 01

    Hess: the July 2025 Chevron transaction delivered 1.025 Chevron shares for each Hess share. The reconstructed value uses public monthly adjusted-price history through the deal, then Chevron's adjusted-price change through December 2025.

  2. 02

    Family Dollar: the July 2015 Dollar Tree transaction delivered US$59.60 in cash plus 0.2484 Dollar Tree shares per Family Dollar share. The stock portion follows Dollar Tree through December 2025; the cash portion compounds at the monthly three-month Treasury-bill rate.

  3. 03

    XL Capital: AXA paid US$57.60 cash per share in September 2018. The reconstructed value takes that cash consideration and compounds it at the monthly three-month Treasury-bill rate through December 2025.

These rows are careful public-source reconstructions, not a licensed total-return database. Their deal terms are primary-source facts; archived entry-price and distribution history can be upgraded later without changing the calculation design.

The money journey

Twenty $1,000 annual cohorts

Think of this as twenty separate envelopes, not one magical compounding account. At the beginning of each trade year, a fresh $1,000 envelope buys that year's selected company and stays invested through the last close of 2025. The 2006 envelope therefore compounds for twenty calendar years, the 2007 envelope for nineteen, and the 2025 envelope for one. Cash distributions are reflected through adjusted closing prices, which is the standard shorthand for reinvesting ordinary dividends. The final total is the sum of every envelope's 2025 ending value.

Planned contributions

US$20,000

US$1,000 at the start of each of 20 trade years.

Selected-company ladder

US$201,413

All 20 cohorts through 2025: 17 direct price histories and 3 clearly marked corporate-action reconstructions.

Corporate-action ledger

3 cohorts

Hess, Family Dollar, and XL Capital are carried through their actual deal consideration instead of being dropped when their tickers ended.

For a like-for-like reference, the full twenty-envelope ladder ends at US$91,245 in SPY and US$166,906 in QQQ. The selected-company total is US$201,413, including three public-source reconstructions. Formula: a direct cohort value = US$1,000 x (2025 adjusted close / entry adjusted close). For a corporate action, the value at the deal is converted into the stock and cash consideration actually received; successor stock is held to the end of 2025 and cash compounds at monthly 3-month Treasury-bill rates. The three reconstructed rows are deliberately labelled in the table so a better archival series can replace them without changing the rest of the study.

Experiment 01

The rule, in plain English

  1. 01

    At each calendar year's close, rank the companies that were in the S&P 500 at that time by their adjusted one-year return.

  2. 02

    At the next year's first available close, place a simulated 100% allocation in the winner and hold through that year's final close.

  3. 03

    Show the same year beside SPY and QQQ, using adjusted closes so the comparison includes ordinary distributions.

  4. 04

    Make no midyear substitutions. When an acquired company ends, carry its actual deal consideration forward: successor shares stay invested, while cash earns the published three-month Treasury-bill path through the final date.

Experiment 01

Twenty contributions, one ending date

Every row begins with US$1,000 at the first available close of its entry year and ends at the final adjusted close of 2025. SPY and QQQ receive the same US$1,000 on the same entry date. This makes the table a contribution-by-contribution audit: early rows have more years to compound, while later rows have less time but exactly the same starting amount.

Entry yearYears heldSelected companyCompany that yearCompany valueSPY that yearSPY valueQQQ that yearQQQ valueData status
200620National Oilwell Varco NOVUnavailableUS$656+15.79%US$7,797+7.09%US$17,374Verified
200719Hess HES -> CVXUnavailableUS$3,676+5.49%US$6,861+18.67%US$16,546Reconstructed corporate action
200818Jacobs Engineering JEC -> JUnavailableUS$1,790-37.00%US$6,571-41.65%US$14,151Verified ticker continuity
200917Family Dollar FDO -> DLTR + cashUnavailableUS$4,015+26.46%US$10,004+54.68%US$22,906Reconstructed corporate action
201016XL Capital XL -> cashUnavailableUS$4,989+15.06%US$8,020+19.21%US$15,228Reconstructed corporate action
201115Netflix NFLX-60.56%US$36,787+2.11%US$7,016+3.40%US$12,663Verified
201214McGraw-Hill MHP -> SPGI-20.59%US$14,120+16.00%US$6,848+18.13%US$12,194Verified ticker continuity
201313Booking Holdings PCLN -> BKNGUnavailableUS$8,526+32.39%US$5,848+36.63%US$10,198Verified ticker continuity
201412Netflix NFLX-7.21%US$18,089+13.69%US$4,577+19.12%US$7,762Verified
201511Skyworks Solutions SWKS+6.64%US$1,075+1.38%US$3,997+8.43%US$6,481Verified
201610Netflix NFLX+8.24%US$8,527+11.96%US$4,003+7.10%US$6,032Verified
20179NVIDIA NVDA+81.99%US$74,237+21.83%US$3,497+32.66%US$5,464Verified
20188Align Technology ALGN-5.74%US$695-4.38%US$2,875-0.13%US$4,084Verified
20197AMD AMD+148.43%US$11,373+31.49%US$3,031+39.12%US$4,144Verified
20206AMD AMD+99.98%US$4,362+18.40%US$2,291+48.63%US$2,945Verified
20215Tesla TSLA+49.76%US$1,849+28.71%US$1,981+27.42%US$2,047Verified
20224Devon Energy DVN+50.86%US$981-18.11%US$1,509-32.58%US$1,568Verified
20233Occidental Petroleum OXY-4.08%US$708+26.29%US$1,863+54.86%US$2,365Verified
20242NVIDIA NVDA+171.25%US$3,874+25.02%US$1,478+25.60%US$1,543Verified
20251Vistra VST+17.66%US$1,084+17.88%US$1,180+20.62%US$1,210Verified

Experiment 01

What the experiment can and cannot say

  • The membership list is point-in-time, but free historical price feeds do not perfectly preserve every delisted company, corporate action, or old ticker. The ledger is therefore a research record, not an institutional-quality index reconstruction.
  • The rule is highly concentrated. A broad index spreads the fate of hundreds of businesses; this experiment accepts the fate of one prior winner.
  • A large historical outcome may be driven by a few unusual years. It does not mean annual winners will continue winning, or that the same path is tolerable in real time.
  • SPY and QQQ are benchmarks, not opponents to defeat. Their diversification, lower turnover, and simpler ownership are part of their value.

Experiment 01

Research sources

S&P 500 historical constituent recordS&P 500 total-return historyAnnual-winner return visual (S&P Dow Jones Indices data)Chevron completion of the Hess acquisitionDollar Tree completion of the Family Dollar acquisitionAXA completion of the XL Group acquisitionFRED: 3-month Treasury bill secondary-market rate
Best Company of the Year | David's Notes